Google blocked 8.3 billion ads in 2025 as Gemini AI transforms enforcement strategy

Google Blocks Record Number of Malicious Ads Using AI

Google blocked 8.3 billion ads globally in 2025, a dramatic increase from 5.1 billion the previous year, according to the company's Ads Safety Report released Thursday1

1

. The surge marks a significant escalation in the platform's battle against deceptive advertising, yet the numbers tell a more nuanced story about how the tech giant is reshaping its approach to platform safety. Despite blocking significantly more content, Google suspended far fewer advertiser accounts than the jump in blocked ads might suggest, signaling a fundamental shift in its enforcement strategy3

3

.

The disparity between blocked content and account suspensions reveals a transition from banning bad actors outright to blocking individual policy-violating ads on a case-by-case basis. Google's Gemini models now power this more granular approach, analyzing billions of signals including account age, behavioral patterns, and campaign characteristics to identify threats before they reach users2

2

.

AI Ad Enforcement Catches 99% of Violations Before Users See Them

Google's AI-driven enforcement systems intercepted more than 99% of policy-violating ads before they were shown to users last year, representing a significant improvement in proactive detection capabilities1

1

. The company attributes this success to its Gemini models, which can detect patterns across large campaigns and understand intent even when malicious content is designed to evade detection2

2

.

Keerat Sharma, VP and general manager of ads privacy and safety at Google, explained during a virtual briefing that the company has shifted toward "more targeted, AI-driven enforcement at a much more granular level, on a creative level, as opposed to using a much more blunt instrument, like advertiser suspensions"1

1

. This refined approach has delivered measurable results, reducing incorrect suspensions by 80% year over year3

3

.

Scammers Using Generative AI Drive Enforcement Escalation

The dramatic increase in blocked ads reflects the growing sophistication of threats as scammers leverage generative AI to produce deceptive ads at industrial scale. Google noted that bad actors are now using generative AI to create convincing advertisements that spoof legitimate businesses, making detection more challenging1

1

. In response, Google is fighting fire with fire, deploying its own AI systems to detect and block deceptive content in real time2

2

.

Among the blocked content, 602 million ads and 4 million advertiser accounts were linked to scams in 20251

1

. By the end of last year, the majority of Responsive Search Ads created in Google Ads were reviewed instantly, with harmful content blocked at submission—a capability Google plans to extend to more ad formats this year2

2

.

Regional Patterns Show Varying Enforcement Challenges

In the U.S., Google removed over 1.7 billion ads and suspended 3.3 million advertiser accounts in 2025, with ad network abuse, misrepresentation, and sexual content ranking among the most common violations1

1

. India, Google's largest market by users, saw the platform block 483.7 million ads—nearly double the previous year—while advertiser account suspensions fell to 1.7 million from 2.9 million3

3

. The primary violations in India involved trademarks, financial services, and copyright issues1

1

.

Layered Defenses and Advertiser Verification Prevent Account Creation

Google has implemented layered defenses including advertiser verification, a process requiring businesses to confirm their identity before running ads, which Sharma said helps prevent bad actors from creating accounts in the first place1

1

. This preventive layer contributes to the decline in suspensions while maintaining platform safety3

3

.

Sharma noted that enforcement numbers are likely to fluctuate over time as Google rolls out new defenses and bad actors adapt their tactics. The company's goal remains intercepting harmful ads as early as possible in their lifecycle, emphasizing the ongoing arms race between platform security and evolving threats1

1

. The shift mirrors Google's broader push to integrate Gemini models more deeply into its core products and infrastructure, where AI increasingly automates campaign creation, detects violations, and responds to emerging threats in real time1

1

.